Autodesk Dreamcatcher
Project Description

This is a one 1-pager I created after I became interested in Autodesk's latest artificial intelligence project called Dreamcatcher. Dreamcatched is an exceptionally advanced design tool that can take in a simple set of parameters and run through thousands of potential designs that fit the requested settings.

01. Brainstorming

I was thinking one day about the applications for 3D printing beyond little trinkets or mechanisms, and one thing I thought was interesting was the ability the technology has to manufacture something that wouldn't be plausible or even possible with traditional techniques. After doing more research, I was really impressed when I found Autodesk's project using AI to create the optimal solution model based on the input constraints.

02. Resources

The project is still very young in the R&D lab at Autodesk, but there have definitely been some major milestones as of recent. At the time of writing, this project has proven successful in multiple use cases, such as in the redesign of a racecar chassis, a drone body, an airplane, and most recently, even an entire office space. In addition to the TedX talk and youtube video from Autodesk themselves, I also had the opportunity to get a sneak-preview of the brand new office before it officially opened in the MaRS building downtown Toronto. Their office was actually created using their generative design software to demonstrate the crazy advancements they'd been making.

03. The Future

The Dreamcatcher project is still super young and is only the first edition of a new era of software tools - as we further develop our understanding of machine learning, software will begin to take a much more active role in the user experience. They won't be correcting our grammar, they'll be writing it. Likewise, they won't be blank slates that we try to cram information into, they'll be activate particpants in every task imaginable, from writing a story to 3D-modeling a new drone chassis.

Interested? Check out the 1-pager here! View PDF